236 Dr. W. Marcet. 



The upper agitator is disposed in such a way as to drive the air of 

 the chamber through a mass of ice roughly broken up, and held in a 

 cylindrical tin vessel open at the top, suspended from the roof of the 

 chamber. The cold air having from its increased density a tendency 

 to fall, is taken up by the lower ventilator and driven upwards ; by 

 such means a circulation of the air in the chamber is maintained 

 through the ice-holder. Should the temperature of the chamber rise 

 during the experiment, by increasing the draught through the ice 

 more ice is melted and the rise is checked ; the reverse holds equally 

 good. 



The ice used for absorbing the heat emitted by the person under 

 experiment delivers its water into a flask holding a thermometer 

 divided into fiftieths of a degree centigrade ; the flask and ther- 

 mometer are both weighed. The flask hangs from a hook on the 

 side of a tube projecting from the bottom of the ice-holder. 



Besides the thermometer in the flask, there are three other ther- 

 mometers all centigrade, also divided into fiftieths of a degree, 

 connected with the calorimeter ; the bulb of one of them projects 

 into the copper chamber while its stem is carried outside above the 

 wooden chamber ; a second has its bulb fastened down to the side of 

 the copper chamber by means of a strip of copper which covers it 

 entirely, its stem also projecting outside ; the third thermometer 

 is passed through the wooden chamber into the annular space, whose 

 temperature it shows during the experiment. 



The successive stages of an experiment are as follows : the heat 

 emitted from the body is first rapidly distributed throughout the 

 chamber, then it is absorbed by the mass of ice, reappearing in 

 measurable form as water. Knowing that 79 calories are required 

 for melting 1 gram of ice, the total calories corresponding to the 

 ice melted are easy to calculate. 



Some of the heat emitted falls upon the brightly polished surface 

 of copper of the chamber, most of it is reflected into the chamber, but 

 a certain proportion becomes absorbed in the metal, and is determined 

 by the thermometer attached to the walls of the chamber. Perhaps 

 a very small amount passes through the copper walls into the 

 annular space, it might have been neglected, but has been taken into 

 account in every experiment. A number of experiments showed 

 that the copper was equally heated in every part, or very nearly so, 

 while the test experiments made with hydrogen gas placed that 

 question quite at rest. 



The thermometers were generally read and the readings recorded 

 every ten minutes, the temperature of the copper being used as a 

 guide towards the maintenance of a constant temperature through- 

 out the instrument. 



It will be readily understood that a difference of as much as 1 



I 



